Development of collision avoidance system for multiple autonomous mobile robots

This article presents a collision avoidance system for multiple robots based on the current autonomous car collision avoidance system. The purpose of the system is to improve the current autonomous car collision avoidance system by including data input of other vehicles’ velocity and positioning via vehicle-to-vehicle communication into the current autonomous car collision avoidance system. There are two TurtleBots used in experimental testing. TurtleBot is used as the robot agent while Google Lightweight Communication and Marshalling is used for inter-robot communication. Additionally, Gazebo software is used to run the simulation. There are two types of collision avoidance system algorithm (collision avoidance system without inter-robot communication and collision avoidance system with inter-robot communication) that are developed and tested in two main road crash scenarios, rear end collision scenario and junction crossing intersection collision scenario. Both algorithms are tested and run both in simulation and experiment setup, each with 10 repetitions for Lead TurtleBot sudden stop, Lead TurtleBot decelerate, Lead TurtleBot slower speed, and straight crossing path conditions. Simulation and experimental results data for each algorithm are recorded and tabulated. A comprehensive comparison of performance between the proposed algorithms is analyzed. The results showed that the proposed system is able to prevent collision between vehicles with an acceptable success rate.